65
(Sacco, 1995). Throughout time, the definition violence became broader and broader (Mucchielli, 2010). The analysis must therefore take account of both of these possible effects in its selection of data in the present. There are no statistics that speak to violent crime to draw comparisons from history, but it can be subjectively analysed that violent crimes are increasing in terms of reporting.
66
availability of more coders, analysis had to be done alone. Despite these drawbacks however, discourse analysis presents a number of advantages. A range of topics can be explored without limitations. The researcher also has complete control of the research material and therefore can focus on their topic of interest (Edley, 2001).
Discourse analysis also bears linguistic abilities to analysis since it is rooted in linguistic philosophy (Potter & Wetherell, 2004). The articles chosen were approached from different syntactic and semantic perspectives. Subjectivity could not be avoided in linguistic analysis because the type of issues dealt with here cannot reach definitive truths (Thompson, Gray &
Kim, 2014). Analysts tend to disagree on what words signify bias or violence or other lexical effects (Westberry & Franken, 2013). Nevertheless, objectivity was strived for in the analysis process.
4.6.1. Data Analysis Process
What the researcher found as interesting in the articles was highlighted and considered for analysis. The program Nvivo 10 for Windows was used in this process. Nvivo 10 is a platform for analysing all forms of unstructured date. With this tool, data was quickly interrogated using the search, query and visualisation tools embedded in the program. Though the program is a time saver, the researcher is still in charge of manually coding, analysing and interpreting the data. The program merely serves as an elaborate and modern organisational tool. Articles had to be reread a number of times to ensure familiarity with the data and to ensure that nothing of value be missed.
Themes were then coded in the program. The themes were drawn from the research questions to form three main themes perpetrators, victims and criminal justice system. The information that went under these themes was then painstakingly gleaned from the articles, as well as other themes that were of interest. Parker (2008) states that discourse is not just a collection of observable statements but it rather operates through words and imagery which communicates what is not communicated everywhere else. Discourse then becomes the 'language' use for communicating social practices from certain viewpoints. With that in mind, words and imagery that could be coded as discourse were identified and classified as either main discourse or sub- discourse. Once the themes and discourses were categorised they were then analysed.
After reading the text to get a general overview of how violent crime is portrayed, the steps for discourse analysis proposed by Ian Parker (1992) were followed. Table 4.3 presents these
67
steps. These are the same steps that were followed by McRitchie and Seedat (2008) in their analysis. The same steps were followed because of the similarities between the current study and that of the two researchers.
Each article was individually analysed and the subjects (e.g., he, she, they, suspect, victim, police, etc.), objects (e.g., guns, neighbours, mother, girl, corpse, etc.) and institutions (courts, police) were identified. Both dominant and non-dominant discourses were identified and recorded. To identify these discourses, terminologies and phrases as well as headline messages were noted. What did not appear in the data was also noted to identify missing discourses. After giving a general picture of violent crime as presented by the identified discourses, illustrative texts of the discourses were highlighted. The final step included the integration of the identified discourses and illustrative texts into a critically oriented interpretive framework that factored in the news values proposed by Jewkes (2010). The headlines of the articles were also analysed for key messages.
Table 4.3: Steps for Discourse Analysis (Parker (1992) as cited by McRitchie & Seedat, 2008) A discourse is realised in texts:
1. Treating the objects of study as texts which are described, put into words.
2. Exploring connotations through some sort of free association.
A discourse is about objects:
3. Asking what objects are referred to, and describing them, (i.e. itemise the objects).
4. Talking about the talk as if it were an object, a discourse.
A discourse contains subjects:
5. Specifying what types of person are talked about in this discourse, some of which may already have been identified as objects, (i.e. itemising the subject).
6. Speculating about what they can say in the discourse, what you could say if you identified with them (what rights to speak in that way of speaking). What they might say within the framework of rules presupposed by the next.
A discourse is a coherent system of meanings:
7. Mapping a picture of the world this discourse presents.
8. Working out how a text using this discourse would deal with objections to the terminology.
68 A discourse refers to other discourses:
9. Setting contrasting ways of speaking, discourses, against each other and looking at the different objects they constitute.
10. Identifying points where they overlap, where they constitute what look like the same objects in different ways.
A discourse reflects on its own way of speaking:
11. Referring to other texts to elaborate the discourse as it occurs, perhaps implicitly, and address different audiences. How these ways of seeing or speaking address different audiences.
12. Reflecting on the term used to describe the discourse, a matter which involves moral/political choices on the part of the analyst. Labeling the discourse.
A discourse is historically located:
13. Looking at how and where the discourses emerged.
14. Describing how they have changed, and told a story, usually about how they refer to things which were always there to be discovered.
Discourses support institutions:
15. Identifying institutions which are reinforced when this or that discourse is used. Institutions which are supported by the discourse.
16. Identifying institutions that are attacked or subverted when this or that discourse appears.
Discourses reproduce power relations:
17. Looking at which categories of person gain and lose from the employment of the discourse.
Who are the beneficiaries?
18. Looking at who would want to promote and who would want to dissolve the discourse.
Discourses have ideological effects:
19. Showing how a discourse connects with other discourses which sanction oppression.
20. Showing how the discourses allow dominant groups to tell their narratives about the past in order to justify the present, and prevent those who use subjugated discourses from making history.
69